Theory & Science (2004)

ISSN: 1527-5558

Misreading Schutz: A Response to Dennis on ‘Lynch on Schutz on Science’

Michael Lynch
Department of Science & Technology Studies
Cornell University
Ithaca, NY

I am flattered that Alex Dennis (2004) has devoted an entire article to criticizing my work, and I’m grateful that this journal has invited me to reply. Dennis’s criticisms require me to dust off some work that I wrote some time ago. The book on which he focuses most of his criticism was published more than ten years ago (Lynch, 1993), and the paper on Schutz and sociology of science was published five years before that (Lynch, 1988). Fortunately, my book is still in print, and the issues Dennis raises are, for the most part, time-insensitive.

In this brief response, I will avoid giving a direct point-by-point rebuttal to many of Dennis’ arguments and interpretations. One reason I am backing away from an invitation to extend the chain of scholarly readings – my response to Dennis’ reading of my reading of Garfinkel’s reading of Schutz – is that I find it very difficult to locate exactly what all this reading and rereading is about. Dennis often paraphrases and characterizes the texts he characterizes, without giving precise references to textual sources. His paraphrases of my arguments often impute schematic distinctions, stark dichotomies, and other categorical structures that I tried to avoid, or to at least qualify, when I originally wrote the texts in question. Of course, his criticisms still may have purchase, despite my original intentions (or what I now say my intentions were), but I cannot write a sensible response to his criticisms when I cannot locate where I said what he imputes to me. Consequently, I will not defend much of what Dennis’ paper says that I argued, because as far as I am concerned, I did not say just that; or, at least, I would hope I didn’t. Instead, I will use his characterizations as an occasion to clarify some points from the writings in question, in hopes that readers may find something of some value in them. 1

The Natural/Social and Topic/Resource Distinctions

To begin, take a brief statement on page [8 ms. copy] of Dennis’ article, where he says, ‘Schutz however, Lynch argues, advances a view of scientific practice which is incapable of being empirically tested, and which has been superseded by subsequent ethnographic investigations of what scientists actually do’. He does not reference where I supposedly argue this, but perhaps he has in mind passages in which I say that ‘Schutz has been accused of exempting natural science research from a thoroughgoing ethnography of practical actions and practical relations’ (1993: 135), or in which I suggest that ‘[i]t may be fair to say that Garfinkel’s and Cicourel’s reliance on Schutz was part of a “protoethnomethodological” development that has been superseded by contemporary research’ (1993: 141). Dennis’ characterization misses the point of what I argued. It also passes over qualifiers in my statements, and other indications in the writing that surrounds them, which suggest that I do not fully endorse the accusation of Schutz I mention, and that I do not subscribe to the idea that ‘protoethnomethodological’ work that relied upon Schutz simply has been superseded by later research in ethnomethodology and conversation analysis.

Contrary to what some of Dennis’ formulations suggest, I did not envision a linear scheme of stages – proto-ethnomethodology, ‘mature’ ethnomethodology, post-analytic ethnomethodology – through which a unitary field of ethnomethodology would progress. Nor did my terms suggest that proto-ethnomethodology is ‘flawed’ or marked by some sort of mistake. Far from offering ‘post-analytic ethnomethodology’ as a corrective, I suggest (1993: 152-3) that it may be impossible to proceed with a social science (or something like a social science) that does not rely upon some kind of principled (‘protoethnomethodological’) distinction between ordinary and scientific (or academic) analysis.

The point about Schutz [being accused of] exempting natural science alludes to criticisms made by sociologists of scientific knowledge (especially Bloor, 1976; but also Latour & Woolgar, 1986[1979]) that the sociology of knowledge had exempted natural scientific practices from its purview. Such criticism was mainly directed against Karl Mannheim’s sociology of knowledge, but as I suggested, it implicated Schutz’s conception of the world of scientific theorizing as a cognitive ‘attitude’ that transcends the mundane, everyday subject matter of sociology. The issue, as I construed it, was not that Schutz put forward an incorrect description of actual scientific practice, or that he described natural science in a way that defied empirical verification or falsification, but that his picture of natural science was part of a logic for a social science that sets up a principled distinction between the substantive topics of social research and the analytical resources through which such topics are investigated (Lynch, 1993: 138). This distinction became familiar as the topic-resource distinction, as explicated in Garfinkel and Sacks (1970) and Zimmerman & Pollner (1970), and later incorporated into the sociology of scientific knowledge through the ‘discourse analysis’ of Gilbert and Mulkay (1984). Although the distinction was valuable for distinguishing an orientation to social practices from a pragmatic engagement in them, it can be confused with a conception of objective analysis that promotes itself as higher-level, purer, more reflexive, or more correct than the topicalized actions and reasonings. In other words, it can be confused with one or another variant of objectivism that turns social ‘actors’ into unconscious, falsely-conscious, and unreflexive inhabitants of a sociological model. In other words, variants of constructive analysis that Garfinkel (1967) counterpoised to ethnomethodology can return, and frequently do return, to ethnomethodology under the guise of the topic/resource distinction.

As I acknowledge, Schutz said little about the natural sciences, and what he did say was incidental to his discussion of the social sciences and their relations to commonsense knowledge in, of, and as the lifeworld. In his writings on rationality and multiple realities, Schutz (1962a, b) occasionally characterized a contemplative ‘attitude’ or ‘world’ of scientific theorizing, which he likened to a dream world or world of the theatre, which enjoys a distinctive cognitive style and is bounded from the everyday world of wide-awake practical engagements. Schutz recognized that science also has its workaday aspects – akin to the day to day ‘laboratory life’ later documented in workplace ethnographies – and so the cognitive attitude he characterized was separate from everyday scientific practice, just as it transcended the broader domain of practical, common sense reasoning. I invite readers to examine the relevant texts by Schutz, and to see for themselves if he does not present a picture of the ‘world’ of scientific theorizing as, predominantly, a cognitive attitude that transcends the day-to-day engagements and interactions of scientific work. This ‘attitude’ selectively substitutes a unified corpus of scientific knowledge and a set of explicit methodological rules, for the taken-for-granted presuppositions and practices of common sense. The key issue I raise about this picture of a unified science, demarcated from common sense is not that it is ‘incapable of being empirically tested’, but that it is consistent with a classic picture of natural science that was extant in Schutz’s day and which has undergone sustained criticism in recent decades by philosophers, historians and sociologists of science. Empirical studies of laboratories were inspired by, and served further to document, a contrastive picture of science. Such criticism need not be accepted wholesale to support the point that Schutz’s picture of natural science has been attacked (and arguably displaced) with an alternative picture that emphasizes epistemic discontinuity and disunity. Moreover, I argued that Schutz sets up his program for a social science with a classic picture of natural science. The point about Schutz is not that he was wrong, or incapable of being proved wrong, in what he said about the natural sciences, but that his seminal writings include conceptions of scientific methodology that were uncritically incorporated into some versions of ethnomethodology’s programme.

Post-Analytic Ethnomethodology &Amp; Proto-Ethnomethodology

The radical promise of ethnomethodology, at least as I view it, is to undertake research without initially setting up a secure refuge for ‘analysis’ that would be uncontaminated by the commonsense knowledge and everyday methods studied. Contrary to Dennis’ characterizations, I did not suggest that ethnomethodology had progressed, or would progress, through a series of stages – proto-ethnomethodology, Garfinkel’s ‘mature’ ethnomethodology, and ‘post-analytic ethnomethodology’. Instead, what I dubbed ‘post-analytical ethnomethodology’ is, and has long been, a possibility or tendency – more-or-less manifest, sometimes obscure, often co-existing with contradictory tendencies – in ethnomethodological investigations. As a practical and professional matter, it is often necessary to suppress or completely hide that tendency, and to put forward ethnomethodological ‘findings’ couched in the familiar analytical idioms of a social science. But, as Garfinkel occasionally suggests, there is a deeply subversive potential to ethnomethodology, and it is that potential that I aimed to encourage. Similarly, when I spoke of ‘proto-ethnomethodology’, I did not mean to refer to an early ‘stage’ in a progression toward a later, and better, research program. Instead, I wrote of an always-available possibility for securing the form and credibility of one’s own research by setting up a difference between a community of analysts and the communities of practice under analysis.

Recovering Analysis

Conversation(al) analysis (CA) is a case in point. CA developed an original conception of analysis that distinguished its methodology from that of a more standard mode of sociolinguistic analysis. Harvey Sacks’ (1992) lectures, and many of the early papers by Sacks, Emanuel Schegloff, Gail Jefferson, and their students made abundantly clear that names for syntactic, semantic, and other structural linguistic units were to be viewed, in the first instance, as vernacular expressions. Accordingly, linguistic and pragmatic categories such as ‘sentence’, ‘statement’, ‘question’, or even ‘word’, feature in an embodied analysis that is endogenous to, and constitutive of, an ordinary conversation over the course of its production. There is no need for professional grammarians to direct or correct such endogenous analysis, and the CA practitioner’s aim for the most part is to explicate the intricate workings of the analysis embedded in, and productive for, specific occasions of social interaction. CA developed its own characteristic research procedures for tape recording, transcribing, identifying, classifying, and collecting ‘data’, and the field quickly became, and remained, empiricist and technical in form. While such development was a mark of its success, it also lived in tension with the original ethnomethodological sense of analysis as a thematic feature of a constitutive methodology. 2 As I have argued elsewhere, such ‘loss’ is a matter of degree, and it remains possible to recover (or, perhaps restore) an ethnomethodological understanding of some of the more technical devices in the CA instrumentarium. For example, there should be no question that ‘turn-transition-relevance place’ 3 and ‘next-turn-repair-initiator’ 4 are technical terms in CA’s analytical vocabulary. Unlike ‘sentence’ or ‘proposition’ they are not vernacular words that linguists and analytical philosophers use as names for technical phenomena. However, contrary to Schegloff (1984) – who draws a hard distinction between vernacular categories and analytical features of talk-in-interaction – the point I would like to make about such terms as ‘turn-transition-relevance-place’ is not that they remain aloof from commonsense intuitions. Instead, in my view, ‘turn-transition-relevance-place’ is an insightful way of speaking about a ‘place-time’ that is located entirely in, and through, vernacular intuitions. It is a ‘place-time’, for example, that depends for its ‘existence’ on a locally competent recognition – a recognition that presumes that relevant others also recognize – that a ‘question’ has been asked, or that it has just now been ‘answered’; it is a ‘place-time’ that arises when a participant in a round of jokes ‘finds’ that it is time to start (or stop) laughing, or to tell another joke; it is a ‘place-time’ that arises when a recipient recognizes that a story is finished and it is time to show appreciation; it is a ‘place-time’ that informs a speaker’s impression that her interlocutor ‘wasn’t listening’. The fact that an evident response by another participant in a conversation marks the existent and normative ‘nature’ of the ‘place’ the preceded and motivated that response has been turned into an ontological basis for grounding or testing the professional conversation analytical inferences. The response becomes a ‘proof criterion’ (Sacks et al. 1974: 728-9) for testing the adequacy of analytical inferences, and conversation analysts cite such a criterion as a basis for distinguishing CA from modes of analysis that rely upon vernacular intuitions. This method of verification becomes the means for distinguishing ‘analysis’ – now a possession of a community of professional analysts – from mere vernacular intuitions (what any competent member can understand of an utterance when appropriately placed in the relevant situation). The problem with proof procedure, as I have argued previously, is that it is founded in a claimed recognition of what the participants studied do, and understand that they are doing, in any given instance. It does not verify what was done in a way that reverts to a professional community’s technical understandings. Indeed, I would argue, and have argued (Lynch 2000), that this ‘technical’ term – turn-transition-relevance place – invites consideration of whether it makes any sense to speak of an analysis of such things that is grounded in a professional community’s distinctive techniques and technical vocabularies. Accordingly, the idea of ‘post-analytical’ ethnomethodology would divest itself of any aim to replace vernacular understandings with professional analysis, and would seek instead to demonstrate how vernacular understandings are endogenous to the organization of everyday and (for those inclined to investigate them) specialized activities.

Cognitive Sociology Respecified

In the case of CA, a ‘post-analytic’ ethnomethodology would not simply occur ‘after’ CA completed its work, nor would it attempt to deliver a knockout blow to the existing practice of CA. On the one hand it would revert to a conception of ‘analysis’ that was, and is still, explicitly part of CA’s pedagogy and practice. On the other hand, it would attempt to grasp, elucidate, and develop the radical import of that conception. To clarify further the ‘post-analytical’ approach I intended to encourage, allow me to speak of another example of a research program with somewhat ambiguous footing in ethnomethodology: ‘cognitive sociology’. Aaron Cicourel’s (1974) programmatic writings on cognitive sociology draw from Schutz, as well as from ethnomethodology, cognitive psychology and linguistic anthropology. A simple (though, in my view, true) understanding of Cicourel’s project is that it applies themes from the cognitive sciences to the analysis of socially organized, institutionally located conduct. In this respect, Cicourel anticipates later developments in the analysis of ‘distributed cognition’ or of cognitive activities in everyday social settings (Neisser 1982; Hutchens 1995). However promising and successful, such an approach differs from one that respecifies the very idea of ‘cognition’ (Coulter 1991). The difference is not exhausted by the fact that Cicourel, for the most part, uncritically reverts to cognitivist assumptions, whereas Coulter critically examines and largely rejects those assumptions. More significant in the context of the present discussion is that Coulter indicates how selected terms for activities that are often (misleadingly) included under the rubrics of ‘cognition’ – terms such as ‘remembering’ and ‘forgetting’, ‘seeing’ and other ‘perceptual verbs’, and ‘mind’ – are in the first (and even last) instance ordinary words used in varieties of contexts of activity (Coulter & Parsons, 1991; Lynch & Bogen, 1996). Just as I recommend that ‘turn-transition-relevance place’ should not be understood as a name for a stable, structurally defined location, Coulter recommends that ‘cognitive’ terms should not be taken as names for stable mental processes in a mind-brain. The import of such respecification is not simply a redefinition of the concept of ‘cognition’. A ‘post-analytic ethnomethodology’ would undertake empirical research to respecify concepts as perspicuous phenomena in specific settings of practical action (Garfinkel, 1991).

How would one go about such a study? Dennis reformulates an outline of some suggestions I made on how to conduct investigations of ‘epistopics’: recurrent themes in philosophy and social theory that tend to be addressed as unitary epistemological concepts, but which ethnomethodology treats as situated phenomena. 5 In brief, the recommendation is to locate and explicate specific (often recurrent) occasions on which some topic or set of topics of interest are elucidated. The aim is not to apply conceptual or methodological resources drawn from philosophy and the social sciences, in order to analyze specific events and phenomena, but to place oneself in a position to explicate an ‘analysis’ that constitutes the phenomena in question. For example, remembering and forgetting are phenomena that early psychologists and latter-day cognitive scientists have taken as topics of theoretical and empirical interest. Psychology and neurology have plenty to say about memory, recall, forgetting, and the like, and it may seem reasonable to assume that the technical findings of psychology and neurology should be important for understanding human action and interaction. Accordingly, it may seem sensible to conduct ethnomethodological investigations that show or suggest how ‘cognitive’ phenomena manifest in overt interaction. Whether or not such investigations would rival or supplement experimental psychology would be an open question. However, for at least two reasons ethnomethodological investigations of ‘memory’ – or, more precisely, of specific occasions of recall or non-recall – hold the potential to develop in a radically different direction. First, as a number of ethnomethodologists have argued, often by reference to Wittgenstein’s later writings, there is reason to believe that ‘cognition’ is not a coherent domain and that many of the linguistic and interactional activities that are now included in that domain can be explicated without tracing them back to a deep, unitary, and individual cognitive and/or neurological center (Button et al., 1995). Second, investigation of such activities as activities is likely to reveal circumstantial connections that ‘explain’ their organization in a way that cognitive scientists would find irrelevant. So, for example, it is highly relevant for a judge, juror, or interrogating attorney to ‘analyze’ a witnesses failure to recall by considering the relevance of just how the interrogator’s question was phrased, how it related to prior questions, how the failure to recall may be related to an emergent narrative, how the witness may be positioned in the narrative, and so on and so forth. Moreover, while it may be clear that commonplace judgments about individual psychology are part of the story – for example, when a witness complains that he can’t be expected to remember what he had for breakfast six months ago – an auditor’s sense of such a claim may turn on such matters as the salience of the event in question and considerations such as that the claimant is a high government official whose office keeps files on the very matter he professes not to recall (Lynch & Bogen, 1996: 187).

Following my summary outline for investigating ‘epistopics’, and guided by some existing studies, a ‘post-analytic’ ethnomethodology would locate settings in which ‘recall’ becomes perspicuous, not as an ‘applied’ instance of cognitive science, but for an explication of an endogenous activity that takes account of a broad range of immediate and contextual matters, only some of which have any relation to a ‘cognitive’ domain. Indeed, a cognitive scientist can fairly complain that such an investigation has no obvious, interesting, or useful relation to the aims of a cognitive science. 6 It would be altogether different. And that is the point.

Although I outlined how one might start to develop a ‘post-analytic ethnomethodology’, I also recognized that it is not for everyone, that it has a subversive relation to the social sciences, and that many who would undertake such a study might run into intractable problems with sustaining a career. I also noted that there are good professional reasons for ‘proto-ethnomethodology’: for presenting ethnomethodology as a social science discipline that studies ‘ethnomethods’ and develops analyses that go beyond a naturalistic, or ordinary, grasp of the practices through which members constitute social orders, and which thereby contributes specialized knowledge to sociology, linguistics, cognitive science, and other disciplines. However, as I noted, the professionalized form of ethnomethodology becomes difficult to sustain in the face of the insight that available positions for ‘experts’, ‘analysts’, and ‘methodologists’ are already occupied in and around the communities that produce such methods. To follow the subversive insight to its logical limit would be to enter what I called ‘ground zero’ – ‘an affirmation of the organized and intelligible character of a social world untouched by academic hands’ (1993: 152) – and to say the least this would be difficult ‘ground’ to occupy for anyone seeking academic employment. Any actual attempt to do ethnomethodology as part of a professional career inevitably involves a creative tension between a radical ‘post-analytic’ tendency, which has been part of the field from the outset, and an accommodative ‘proto-ethnomethology’ that finds connections to established theoretical traditions and social science (sub)disciplines. One might describe this as a creative tension rather than a series of stages; a creative tension that becomes lost when accommodation overwhelms the very insight upon which ethnomethodology gets its life and seeks its death. Consequently, a linear series of stages, such the one that Dennis [ms. p. 9] outlines – starting with ‘protoethnomethodology’, passing through ‘ethnomethodology proper’, and then on to ‘postanalytic ethnomethodology’ – would be a march toward collective nihilism and individual unemployment. Fortunately, such a ‘program’ has attracted very few followers.

Misreading the Masters

Aside from any linear progression, I do distinguish Garfinkel’s and his students’ studies of work in the sciences from earlier ethnomethodological writings that trade heavily upon themes borrowed from Schutz, but one reason for drawing this distinction is that Garfinkel himself drew something like it. In lectures and tutorials to his students in the 1970s and early ‘80s, Garfinkel repeatedly announced a new programme in ethnomethodological studies of work (Garfinkel 1986). One trademark ‘policy’ he associated with that program was its studious indifference to scholastic debates about, among others, the writings of Alfred Schutz (Garfinkel 1977). He recommended instead that, when undertaking studies of work in the sciences and professions – work that develops its own genealogies, methodological maxims, training regimens, and bodies of literature 7 – it may be advisable to ‘misread’ the writings of the phenomenologists and social theorists. He was not advising us not to read them; indeed, Garfinkel consistently expressed fascination, and even love, for the brilliant thematic formulations, illuminating passages, and perspicuous examples found in the writings of Husserl, Heidegger, Merleau-Ponty, Gurwitsch and, of course, Schutz. As I understand it, the advice to ‘misread’ these philosophers is an invitation to engage in a reading of a different kind: not only to read Garfinkel’s texts in relation to those of Schutz; or Schutz’s texts in relation to those of Husserl or Heidegger (although as part of the academic life we inevitably will and must do, even enjoy doing, such textual work); but also to abandon responsibility for ‘correct’ literary interpretation, while playfully appropriating gestalt and phenomenological themes (such as ‘topical contexture’, ‘figure-ground’, ‘modalities’ of perception, and ‘mathematization of nature’) and re-discovering them as (sometimes humble, everyday; sometimes esoteric) locally organized performances and productions. At the end of the day there would always be another day: there was no promise of a final theory of society-in-general; no stable glossary of key concepts and definitions. Instead, there would be endless cases through which to re-specify what Schutz, Husserl, or even Garfinkel, most likely did not have in mind, but might as well have meant, in their seminal writings. 8

As I noted earlier, in my response to Dennis I have attempted to avoid the invitation to consider whether I read Schutz, Garfinkel, and the relation between the two correctly. This does not mean that I am indifferent to reading. Reading Schutz, Garfinkel, Husserl, Heidegger, and Wittgenstein has been critical for my own intellectual development. Further, despite their turn away from ‘literary’ approaches to social analysis, Sacks (1992) and Garfinkel (1967) express a deep, if cryptically expressed, engagement with scholarship. However, I believe that the best ‘test’ of how an ethnomethodologist reads Schutz or any other seminal figure should be found, not in a reading of the original texts but in a misreading of those texts in connection with original studies (again, this is not a matter of ‘application’). Consequently, while I am not persuaded by Dennis’ paper that I read Schutz or Garfinkel incorrectly, that is beside the point, and I leave it to others to judge whether I misread them in an interesting way.


Bloor, D. (1976). Knowledge and Social Imagery. London: Routledge & Kegan Paul.

Button, G., Coulter, J., Lee, J.R.E., and Sharrock, W. (1995). Computers, Minds and Conduct. London: Polity.

Cicourel, A. V. (1974). Cognitive Sociology: Language and Meaning in Social Interaction. New York: Free Press.

Coulter, J. (1983). Contingent and A Priori Structures in Sequential Analysis. Human Studies 6: 361-376.

Coulter, J. (1991). Cognition: Cognition in an Ethnomethodological Mode. In Ethnomethodology and the Human Sciences. Button, G. (ed.). Cambridge, UK: Cambridge University Press: 176-95.

Coulter, J., & Parsons, E. D. (1991). The Praxiology of Perception: Visual orientations and Practical Action. Inquiry 33: 251-272.

Dennis, Alex (2004). Lynch on Schutz and Science (this journal).

Durkheim, E. (1982) The Rules of Sociological Method and Slected Texts on Sociology and its Method. Lukes, S. and Halls, W.D. (trans.). New York: Free Press.

Edwards, D., Middleton, D. & Potter, J., (1992). Toward a Discursive Psychology of Remembering. The Psychologist 5: 441-455.

Garfinkel, H. (1967). Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.

Garfinkel, H. (1977). When is Phenomenology Sociological? The Annals of Phenomenological Sociology 1: 2-16.

Garfinkel, H. (1991). Respecification: Evidence for Locally Produced, Naturally Accountable Phenomena of Order, Logic, Reason, Meaning, Method, etc. In and As Of the Essential Haecceity of Immortal Ordinary Society (I) — An Announcement of Studies. In Ethnomethodology and the Human Sciences. Button, G. (ed.). Cambridge, UK: Cambridge University Press: 10-19.

Garfinkel, H. (1996). An Overview of Ethnomethodology’s Program. Social Psychological Quarterly 59: 5-21.

Garfinkel, H. (2002). Ethnomethodology’s Program: Working Out Durkheim’s Aphorism. Lanham, MD: Rowman & Littlefield.

Garfinkel, H. and Sacks, H. (1970). On Formal Structures of Practical Actions. In Theoretical Sociology: Perspectives and Development. J. C. McKinney & E. A. Tiryakian (eds.). New York: Appleton-Century-Crofts: 337-66.

Gilbert, G. N., & Mulkay, M. (1984). Opening Pandora’s Box: An Analysis of Scientists’ Discourse. Cambridge, UK: Cambridge University Press.

Hutchins, E. (1995). Cognition in the Wild. Cambridge, MA: MIT Press.

Latour, B., & Woolgar, S. (1979). Laboratory Life: The Social Construction of Scientific Facts. London: Sage.

Lynch, M. (1988). Alfred Schutz and the Sociology of Science. In Worldly Phenomenology: The Influence of Alfred Schutz on Human Science. Embree, L. (ed.). Washington, D.C.: Center for Advanced Research in Phenomenology and University Press of America: 71-100.

Lynch, M. (1993). Scientific Practice and Ordinary Action: Ethnomethodology and Social Studies of Science. Cambridge: Cambridge University Press.

Lynch, M. (2000). The Ethnomethodological Foundations of Conversation Analysis. Text 20(4): 517-32.

Lynch, M. & Bogen, D. (1996). The Spectacle of History: Speech, Text, and Memory at the Iran-contra Hearings. Durham, NC: Duke University Press.

Neisser, U. (ed.) (1982). Memory Observed: Remembering in Natural Contexts. San Francisco: W.H. Freeman and Co.

Sacks, H. (1992). Lectures on Conversation. Oxford: Basil Blackwell.

Sacks, H., Schegloff, E. A. & Jefferson, G. (1974). A Simplest Systematics for the Organization of Turn-taking in Conversation. Language 50(4): 696-735.

Sacks, H. and Schegloff, E.A. (1979) Two Preferences for the Organization of Reference to Persons in Conversation and their Interaction. In Everyday Language: Studies in Ethnomethodology. Psathas, G. (ed.). New York: Irvington Press: 15-21.

Schegloff, E. A. (1984). On Some Questions and Ambiguities in Conversation. In Structures of Social Action: Studies in Conversation Analysis. Atkinson, J.M. & Heritage, J.C. (eds.). Cambridge, UK: Cambridge University Press: 28-52.

Schegloff, E. A. (1992). To Searle on Conversation: A Note in Return. In (On) Searle on Conversation. Searle, J. R., Parret, H., and Verschueren, J. (eds.). Amsterdam & Philadelphia: John Benjamins: 113-28.

Schutz, A. (1962a). On Multiple Realities. In Collected Papers, Vol. 1. The Hague: Martinus Nijhoff: 207-59.

Schutz, A. (1962b). Concept and theory formation in the social sciences. In Collected Papers, Vol. 1. The Hague: Martinus Nijhoff: 48-66.

Schutz, A. (1964). The Problem of Rationality in the Social World. Collected Papers, Vol. 2. The Hague: Martinus Nijhoff, 1964: 64-90.

Searle, J. (1992). Conversation. In (On) Searle on Conversation. Searle, J. R., Parret, H., and Verschueren, J. (eds.). Amsterdam & Philadelphia: John Benjamins: 7-30.

Turner, R. (1975). Speech Act Theory and Natural Language Use. Pragmatics Microfiche 1(1).

Zimmerman, D. and Pollner, D. (1970) The Everyday World as a Phenomenon. In Understanding Everyday Life: Toward the Reconstruction of Sociological Knowledge. Douglas, J.D. (ed.). Chicago: Aldine Publishing Co.: 80-103.


1. For a quick study of the relevant writings, see Lynch (1993: 133-58; 299-308). The most pertinent of Schutz’s essay for this argument are Schutz (1962a,b; 1964).

2. I am leaving aside an issue that Jeff Coulter (1983) has raised, which is that ‘analysis’ is one of many conceptual rubrics for describing constitutive organizations in, of, and as conversation. When turned into a master concept for the various ways in which conversationalists deploy grammar as ubiquitous resources and occasioned topics, ‘analysis’ tends to connote a picture of what conversationalists are doing that is, perhaps, skewed in the direction of an academic undertaking.

3. Simply, a turn-transition-relevance place is a juncture in an ongoing utterance that is locally recognizable as being a relevant place-time for someone else to begin a ‘turn’ (Sacks, et al. 1974). Some turn-transition-relevance places are conventionally marked as endings and/or as points at which another speaker should respond. Others are points that arise at which another speaker can enter into the conversation. There is a large range of possibility and subtlety in this domain, and one point that is made clear in CA is that such ‘places’, though recognizable in situ, and non-random in the sense that they recur at definable ‘joints’ in the grammatical organization of utterances and exchanges, are not identified by specific signals or a priori structures. A turn-transition-relevance place is the kind of ‘thing’ that distinguishes CA from speech act theory: it’s ‘presence’ is existential, reverting to situated understandings of co-participants in a conversation; as opposed to being a ‘point’ marked by the end of the sentence or a signal such as the rising intonation at the end of a question. It is indicative of that difference that John Searle grossly misunderstood the treatment of turn transition in CA in a brief and critical article he wrote on the subject (Searle, 1992). Searle described CA’s account of turn transition to be a matter of a speaker ending a sentence, and by ending the sentence signaling to other(s) that it was time for someone else to speak, and as Schegloff (1992) points out, such a view misses the irreducibly contingent and interactional way in which such a ‘place’ is produced. Also see Turner (1974) and Schegloff (1984) on the difference between a conversational analytic and speech-act theoretic approaches to interactional phenomena.

4. A ‘next-turn-repair initiator’ is a response that enables the interlocutor to recognize that something is awry in what she just said; the recipient does not correct the target utterance, but instead provides an opportunity to self-repair. Certain uses of the token ‘What?’ enable the speaker of the just-prior utterance to recognize and ‘repair’ some aspect of what they have just said. This contrasts with a more direct correction of the target utterance by the recipient. The way this interactional ‘mechanism’ relies upon a speaker’s recognition of just what might be ‘repairable’ in the prior utterance is a wonderful exhibit of intersubjective attunement.

5. Dennis [draft pp. 9-10] translates my recommendations into a normative program: a series of steps about what a researcher ‘should’ or ‘must’ do. For comparison, see the summary recommendations in Lynch (1993: 300-08). My conception of ‘epistopics’ is cast in relation to studies and debates in social studies of science, and thus it is selectively oriented to topics associated with knowledge, observation, measurement, visual representation, and the like. Garfinkel (1991) provides a list of topics of order drawn from social theory that he respecifies as ethnomethodological phenomena of order to be investigated.

6. For example, some of the cognitive psychologists who responded to a polemical article in The Psychologist advocating a ‘discursive’ approach to remembering (Edwards et al. 1992) seemed puzzled that studies of textual and conversational discourse could have any real bearing on the phenomenon of memory, as studied in cognitive science.

7. The sciences and (academic and clinical) professions are not alone in this regard. The literatures on such subjects as cookery, diet and recreational angling are immense, ancient and varied, including rival doctrines, and a variegated mixture of how-to recipes, histories, observances, and anecdotes, as well as systematic studies and reflections. Those ‘vulgar’ literatures even share some of the scholastic pretensions of many academic literatures, and academics from a broad range of fields contribute to them.

8. Garfinkel’s most recent writings (1996; 2002) adopt Durkheim’s (1982) Rules as a primary source of (apparent) inspiration. Such adoption may seem odd, given the positivism often attributed to Durkheim – and particularly his Rules. One might imagine, however, that Garfinkel weaves Durkheim’s aphorism that ‘the objective reality of social facts’ is sociology’s fundamental principle, together with Husserl’s injunction ‘to the things themselves’. Things become the pivotal word, although one can argue (as I would argue) that Garfinkel’s ‘things’ are neither Durkheim’s nor Husserl’s – nor are they his.